©2002. Commercial distribution is prohibited.
What is XCOSM?
XCOSM is an X-Windows interface to Computational
Optical Sectioning Microscopy (COSM) algorithms [1-7] for removing
out-of-focus light in 3-D volumes collected plane by plane using either
widefield or confocal fluorescence microscopy. The interface allows users
to view and process multiple 3-D datasets within a simple mouse-driven
environment. The software was written for a unix based system. It is possible
to compile XCOSM on a PC with linux. Information on COSM can be found in
the Tutorial
- Image Restoration for 3-D Microscopy and in
lectures given
by J.-A. Conchello . XCOSM can
be dowloaded for free.
| Raw Image |
Processed Image |
 |
 |
Historical note on XCOSM development:
XCOSM was created at the Biomedical Computer Laboratory, a component of
the Institute for Biomedical Computing at Washington University in
St. Louis, MO. Support for XCOSM development came from the National
Center for Research Resources Biomedical Research Technology Program to
the IBC, Lewis. J. Thomas, PI, under Grant #RR01380, A Resource for Biomedical
Computing. The 3-D microscopy component of this grant was led by J.
G. McNally and José-A.
Conchello. Currently XCOSM is not supported by any grants.
Notes/Updates:
5/17/2002: New XCOSM version 2.5 is now available.
In version 2.5 a correction was made to the units of the
"specimen thickness parameter" in the inputs of the PSF generation;
otherwise version 2.5 has the same functionality as version 2.4.
5/17/2002: Linux Users please see the following corrections and comments.
5/21/2001: New XCOSM version 2.4 is now available.
In version 2.4 two utility programs have been added: 1) program
bleach_cor that corrects bleaching decay in the data and 2) program
preprocess that corrects for camera non-uniform response; otherwise version 2.4
has the same functionality as version 2.3.
5/21/2001: New and correct test images and psf data are available
for downloading. You may use these data
to confirm that your installation of XCOSM is correct by generating the same
results with the ones that are posted.
Linux users or people using PCs having problems with PSF
computations make sure that you select the appropriate parameter for the
Intel CPU precision see notes.
Great news for PC users that have problems compiling XCOSM.
We have added a HOWTO
that includes instructions on how to compile the utility programs, the
program for the PSF generation, and the program for the EM deconvolution
algorithm separate from the X-Window dependent user interface so that they
can be used as command line programs using the Cygwin
tools. A copy of the executable files is also available below
- you will still need to install the Cygwin tools. We ask that you also
download
the source so that a copy of your name and address can be obtained for
future updates.
XCOSM features:
I. Five COSM processing algorithms:
-
Regularized Linear Least Squares (LLS) method by C.
Preza et al [1], a relatively fast one-pass
algorithm.
-
Linear maximum a posteriori (MAP) method by C. Preza et al
[2], a relatively fast one-pass algorithm.
-
Expectation Maximization (EM) by J-A.
Conchello et al [3,4,5], a slow iterative
algorithm which requires much longer processing time but provides higher
quality images.
-
Regularized EM using an intensity penalty by J.-A. Conchello
and J. G. McNally [6].
-
Regularized EM using a roughness penalty by J. Markham and
J.-A. Conchello [7].
Any publications with results obtained with XCOSM should
cite the original papers above which led to the development of the package.
View
images processed with XCOSM algorithms.
II. X-Windows graphical interface
II. X-Windows graphical interface
-
3-D image viewing, showing separate 2-D slices of top down (XY) and side
(XZ) views of each loaded volume.
-
Color, size, and contrast of viewed volumes can be manipulated, and individual
voxel values can be determinated.
-
3-D volumes can be loaded from disk, and processed images can be saved
to disk for later inspection.
-
3-D images can be linked allowing determination of voxel values in all
the linked images at the same time.
-
Environments containing sets of 3-D volumes can be saved to disk and loaded
later for inspection.
III. Other features
-
Point Spread Function (PSF) generation which is needed for all the
algorithms for both widefield [8] and confocal
[3,9]
microscopy.
-
Optical Transfer Functon (OTF) generation which is needed by the
LLS/MAP algorithm.
- Utility programs for preprocessing the data to correct for the
non-uniform response of the CCD camera and for fluorescence photo-bleaching.
-
Image oriented support programs that allow data type conversion, sub-volume
extraction, and byte-swapping of 3-D volumes for compatibility across different
platforms.
Below an XCOSM session is shown. The tool bar on the left shows the
funtionality of XCOSM. Section images from two 3D images are displayed in this
session. A section in the XY plane or top view is shown by the top images
and a section in the XZ plane or side view is shown by the bottom images.
The images on the left show
the raw data volume and the images on the right show the result of processing
using the EM algorithm (1000 iterations). The colorbar shows the pheudo
color scale used to display the images. The raw image was provided by Dr.
William Goldman (Department of Molecular Microbiology, Washington University
School of Medicine, St. Louis).
Downloading information :
Downloading
the XCOSM version 2.5 package (~700 Kbytes)
Note that versions 2.0 and up are very different in appearance
and have several updates from version 1.0 which is described in the manual.
This new version of XCOSM supports confocal microscopes and can create
point spread functions (PSFs) for them--please disregard anything in these
pages that mentions that xcosm cannot support confocal microscopes.
NOTE for NETSCAPE users: Before you
click on the Download link below press the shift button down.
Download
test images and files (~30 Mbyte compressed tar file: testdata.tar.gz).
If you are having problems downloading a large file, you may download
each file separately
from here.
Download
executable files for XCOSM version 2.3 for Windows NT.
Submitting a bug
report.
Known
problems and modifications of XCOSM to date.
or send e-mail to: preza@essrl.wustl.edu.
References:
-
C. Preza, M.I. Miller,
L.J. Thomas Jr., J.G. McNally, "Regularized method for reconstruction
of three-dimensional microscopic objects from optical sections", J.
Opt. Soc. Am. A, Vol. 9, pp 219-228, 1992.
-
C. Preza, M.I. Miller, J.A. Conchello, "Image reconstruction for 3-D
light microscopy with a regularized linear method incorporating a smoothness
prior", IS&T/SPIE proc., R.S. Acharya and D.B. Goldgof,
editors, pp 129-139, 1992.
-
J.-A. Conchello, J.J.
Kim, E.W. Hansen,"Enhanced three-dimensional reconstruction from confocal
scanning microscope
images. II. Depth discrimination versus signal-to-noise ratio in partically
confocal images", Appl. Opt., Vol. 33, No. 17, pp. 3740-3750, 1994.
-
J.-A. Conchello, "Super-resolution and convergence properties of the
expectation-maximization algorithm for maximum-likelihood deconvolution
of incoherent images", J. Opt. Soc. Am. A, Vol 15, (10), pp. 2609-2620,
1998.
-
J.-A. Conchello, "Fluorescence photobleaching correction for expectation
maximization algorithm" in Three-Dimensional microscopy: image acquisition
and processing T. Wilson and C.
J. Cogswell editors, SPIE 2412-21, pp138-146, 1995. (PDF,
Postscript,
and unix-compressed
postscript files available).
-
J.-A. Conchello and J.
G. McNally "Fast regularization technique for expectation maximization
algorithm for computational optical sectioning microscopy" in Three-Dimensional
microscopy: image acquisition and processing C.
J. Cogswell, G. S. Kino, and T. Wilson, editors, SPIE 2655,
pp.199-208, 1996.(PDF,
Postscript,
and unix-compressed
postscript files available).
-
J. Markham
and J.-A. Conchello, "Tradeoffs in regularized maximum-likelihood
image restoration" in 3D Microscopy: Image Acquisition and Processing
IV. C. J. Cogswel, J.-A. Conchello, and T. Wilson, editors, SPIE 2984-18,1997.
(PDF,
Postscript,
and unix-compressed
postscript files available).
-
F.S. Gibson, F. Lanni, "Experimental test of an analytical model of
aberration in an oil-immersion objective lens used in three-dimensional
light microscopy", J. Opt. Soc. Amer. A, Vol. 8, pp. 1601-1613, 1991.
-
J. A. Conchello, J. W. Lichtman, "Theoretical analysis of a rotating-disk
partially confocal focal scanning microscope", Appl. Opt. Vol 33 (4),
pp. 585-596, 1994.
-
J. G.
McNally, T. Karpova, J. Cooper, J.-A.
Conchello, "Three-Dimensional Imaging by Deconvolution Microscopy",
Methods, 19, 373-385, 1999.
-
J. Markham and J.-A. Conchello, "Fast maximum-likelihood
image-restoration algorithms for three-dimensional
fluorescence microscopy," J. Opt. Soc. Am. A, 18(05): 1062-1071, 2001.
-
J. Markham and J.-A. Conchello, "Artefacts in restored images due to
intensity loss in three-dimensional
fluorescence microscopy," Journal of Microscopy, Vol. 204, Pt. 2, pp 93-98, 2001.
-
J.G. McNally, C. Preza, J-A. Conchello, L.J. Thomas, Jr., "Artifacts
in Computational Optical-Sectioning Microscopy", J. Opt. Soc. Am. A,
Vol. 11, pp 1056-1067, 1994.
-
C. Preza, J.M. Ollinger, J.G. McNally, L.J. Thomas, Jr., "Point-Spread
Sensitivity Analysis for Computational Optical-Sectioning Microscopy,"
Micron
and Microscopica Acta, Vol. 23, No 4, pp 501-513,1992.
-
Tutorial - Image
Restoration for 3-D Microscopy , Instructors: J.
G. McNally, J.-A. Conchello,
F.
U. Rosenberger, C. Preza,J.
Markham, April 26-27, IBC, Washington University, St. Louis, MO, 1996.
Last Modified May
17, 2002, by
C. Preza, Washington University Three-Dimensional Microscopy Laboratory, St. Louis,
MO